Half-Squares on 30–10 m: Feed Them at the Corner (or Center), Not the End

Why a 49:1 Can Work on an End-Fed Half-Square — and Why That Still Does Not Make It the Best Feed

What a Half-Square Is — and Why We Like It

A half-square is a half-wavelength horizontal wire with a quarter-wavelength vertical wire hanging from each end. The two verticals act like a pair of in-phase, quarter-wave vertical radiators. The top wire is mainly the phasing section. Properly built, the pattern is bidirectional broadside to the array, with useful low-angle vertically polarized radiation — excellent for DX.

The cleanest way to feed a normal monoband half-square is at one of the top corners. That point is a current-fed point: relatively high current, relatively low voltage, and often close enough to 50 Ω for direct coax feed on the design band.

Classic modeling by L. B. Cebik, W4RNL, shows the practical contrast clearly. A top-corner current feed can be in the practical coax range, while a bottom/end voltage feed can be several thousand ohms with large reactance. In one example, the corner-fed case was around 71 + j3 Ω, while the end-fed case was around 4100 − j4000 Ω.

Corner feed, current-fed: often around 50–70 Ω on the design band; no impedance transformer is normally needed — just a good 1:1 current choke at the feedpoint.

Bottom/end feed, voltage-fed: typically several kΩ, often with a large reactive component; it needs a real matching system, not just a random transformer and hope.

Important correction: the midpoint of the top wire is not normally a near-50 Ω coax feedpoint. In the ideal current distribution of a full-wave half-square, the top corners are current maxima. The top-center region is much closer to a current minimum / voltage maximum. You can experiment with top-center balanced feed and a tuner, but it should not be described as another simple 50 Ω current feed.

“But My 49:1 Shows a Nice SWR…” — That May Be Real, But It Is Not the Whole Story

A 49:1 transformer maps 50 Ω to about 2450 Ω. So a 49:1 is not automatically wrong on a voltage-fed half-square. If the bottom/end feedpoint happens to be around 2.5 kΩ and the reactance has been controlled, a 49:1 can give a believable SWR.

The problem is not that a 49:1 can never match a half-square. The problem is that a good SWR through a 49:1 does not prove that the half-square is being fed cleanly, efficiently, or with the intended pattern.

• The end impedance is not fixed. A bottom-fed half-square may be around 2.5 kΩ, 4 kΩ, 5 kΩ, or higher depending on geometry, wire diameter, height, soil, nearby objects, and the presence or absence of a counterpoise.
• The reactive part often dominates. A feedpoint such as 4100 − j4000 Ω is not solved by choosing 49:1, 64:1, or 81:1 alone. The reactance still has to be dealt with.
• Common-mode current can fake success. If the coax shield becomes part of the antenna, the analyzer may show a comfortable SWR while the feedline radiates, the pattern changes, and shack RF appears.
• The transformer capacitor is not magic. The small capacitor in many EFHW transformers is normally used to improve the high-frequency transformer response. It can also affect the measured match, but it should not be treated as proof of efficiency or inefficiency by itself.

Bottom line: a 49:1 can be plausible for a voltage-fed half-square when the feedpoint is near 2.5 kΩ resistive. It is not a universal half-square transformer, and it is not a substitute for measuring feedpoint impedance, transformer heating, common-mode current, and pattern stability.

The Correct Ratio Range Is Usually Higher Than 17–23:1

The previous idea that a half-square end feed commonly needs only about 17–23:1 is too low for the true bottom/end voltage-fed case. That ratio corresponds to roughly 850–1150 Ω when transforming from 50 Ω. That can occur at some displaced or non-standard feedpoints, or on some non-design-band cases, but it is not the normal expectation for the voltage-fed end of a half-square.

For the bottom/end voltage feed, the more realistic design assumption is several thousand ohms. Once the reactance is tuned out or minimized, the approximate impedance-ratio logic looks like this:

If your half-square end feed is really in the 3–5 kΩ range, then your instinct is correct: a 49:1 is not the mathematical center of the target. A ratio somewhere around 64:1, 81:1, or even 100:1 may be closer — but only after the reactive component is handled.

For example, 5000 Ω resistive transformed by 49:1 becomes about 102 Ω on the coax side, which is still only about a 2:1 SWR before losses and stray effects. That is why a 49:1 may “look fine” even when it is not the optimum ratio. But if the feedpoint is 4100 − j4000 Ω, the remaining reactance is a much bigger problem than the exact transformer ratio.

Voltage-Fed vs Current-Fed — Why the Corner Feed Is Still Better

Voltage-fed bottom/end points excite a voltage maximum and current minimum. They are inherently more sensitive to their surroundings:

• stray capacitance from supports, foliage, wet rope, and nearby metalwork,
• leakage across damp surfaces,
• common-mode current on the coax shield,
• matching-box loss and transformer heating, and
• pattern changes caused by the feedline becoming part of the antenna.

Current-fed top-corner points drive a current maximum in the antenna itself. The feed impedance is low enough that direct coax feed is usually practical on the design band. The voltage stress is lower, the feed is more repeatable, and a proper 1:1 current choke can isolate the coax shield much more effectively.

This is why the top-corner feed is usually the best choice if the goal is a predictable, efficient, monoband half-square.

Height still matters: a top-wire height around 0.20–0.25 λ is a good practical target for the classic half-square shape. Too low increases ground interaction. Much higher installations can change the elevation pattern and may introduce stronger higher-angle lobes.

(A voltage-fed bottom/end installation also needs more attention to high-voltage clearance, weatherproofing, matching-box insulation, and common-mode control.)

Recommended Ways to Feed a 30–10 m Half-Square

Corner-feed, best monoband method

• Feed one top corner of the half-square.
• Connect the coax center conductor to the top horizontal wire side of the break.
• Connect the coax shield to the top of the adjacent vertical leg.
• Add a serious 1:1 current choke right at the feedpoint.
• Route the coax away from the antenna at an angle, not down alongside the radiating vertical.
• Trim the antenna for resonance; expect a practical low-impedance feed on the design band.

Bottom/end voltage feed, valid but more demanding

• Expect a feedpoint in the several-kΩ region, not 50 Ω.
• Choose the transformer ratio from measured impedance, not from habit.
• Use 49:1 only if the transformed impedance and heating are acceptable.
• Consider 64:1, 81:1, or a tuned network if the feedpoint resistance is closer to 3–5 kΩ.
• Use a short, deliberate counterpoise or return reference if needed to control reactance and common-mode current.
• Use excellent insulation and weatherproofing because the feedpoint voltage can be high.

Top-center balanced feed, experimental / tuner-fed

• Do not treat the top-center as another simple 50 Ω feedpoint.
• It is closer to a voltage-fed point in the ideal full-wave current distribution.
• It can be used with balanced line and a tuner, but the result is a tuned-wire system, not the same as the clean corner-fed half-square.

Avoid: using a 49:1 at the top corner. The corner feed is already the low-impedance current-feed point. A 49:1 belongs only in the voltage-feed discussion, not in the normal corner-fed half-square design.

Quick-Start Lengths

Use these as starting cuts. The top wire is approximately ½λ and each vertical is approximately ¼λ, with practical shortening included. Expect final trimming for height, wire diameter, insulation, and nearby objects.

(Trim in the real installation. The half-square is sensitive to geometry, height, and surroundings.)

Take-Aways

• For a clean monoband half-square, feed the top corner. That is the low-impedance current-feed point and usually needs only a 1:1 choke.
• The bottom/end feed is high impedance. Your expectation of several kΩ is more realistic than 600–1400 Ω for the true voltage-fed end.
• A 49:1 is not automatically wrong. It is appropriate near 2450 Ω resistive, and may still give acceptable SWR at somewhat higher impedances.
• For 3–5 kΩ, expect ratios closer to 64:1–100:1. But transformer ratio alone does not solve large reactance.
• Do not call the top-center a 50 Ω current feed. The top corners are the current maxima; the top-center is much closer to a voltage point.
• SWR is not proof of performance. Check transformer heating, common-mode current, feedline radiation, and on-air pattern repeatability.

References & Notes

• L. B. Cebik, W4RNL: SCVs Part 4 — The Open-Ended Cousins; half-square feedpoint modeling and voltage-feed examples.
• Portable-Antennas half-square designer: current-fed top-corner half-square versus voltage-fed end-fed half-square behavior.
• OnAllBands / DX Engineering: EFHW transformer ratios, 49:1 logic, and typical end-fed impedance ranges.
• Practical note: always verify the transformer with suitable high-value non-inductive loads, then verify the antenna installation with a current meter or common-mode clamp, not SWR alone.

Interested in more technical content? Subscribe to our updates for deep-dive RF articles and lab notes.

Questions or experiences to share? Contact RF.Guru.